Approach to Indefinite Semantic Conflicts of Words in Collaborative Editing of Design Documents

摘要

Indefinite semantic conflict of words is another category of the problem of semantic consistency of words, besides definite semantic conflict of words. While the definite cases are well solved in our previous work, there are still no effective solutions to the indefinite semantic conflicts of words, which involve a much more complicated process of semantic discrimination. Based on WordNet and our ontology description language FLoDL, we propose a document type based word frequency accumulation algorithm to solve the problems. This algorithm is carried out in two phases based on different document types: a training phase and a discrimination phase. The discrimination phase is mainly based on the words'' frequency information collected in the training process. Finally, through some detailed experiments, a compared analysis is given to show that our algorithm has a far better performance than other previous solutions, thus improves the overall performance collaborative editing